Smallest.ai announced the round on July 30, 2026. Seligman Ventures led, joined by Sierra Ventures, 3one4 Capital, Better Capital, Upsparks Capital, Schema Ventures, Tiny VC, DeVC and Mission Street Capital, among others. The company previously raised an 8 million dollar seed round in October 2025. Founder and CEO Sudarshan Kamath leads a team of nearly 60 people. The product line spans Hydra, a speech to speech model with an asynchronous architecture, Pulse STT Pro, a speech to text model supporting 38 languages with speaker diarization and PII redaction, and Lightning V3.1, a text to speech model.
Traction: six named customers within two years of founding
Smallest.ai names RingCentral, Truecaller, Readymode, Piramal, Kogta and Pocket as customers, a notable list for a company under two years old. The announcement cites customer reported results of up to 80 percent reduction in support costs and tenfold productivity gains. These are company claims and not independently verified. Kamath told TechCrunch the goal is voice agents indistinguishable from humans, with the model handing off complex queries to larger language models only when needed, which keeps response latency close to zero for questions inside the knowledge base.
Market development: voice AI consolidates around latency and cost
Smallest.ai competes with ElevenLabs, Cartesia and Sarvam in a market the company, citing external projections, expects to grow from 2.4 billion dollars in 2024 to 47.5 billion dollars by 2034. The competitive frontier is shifting from voice quality, now largely commoditized, to conversational latency and per minute economics at enterprise scale. Small specialized models are an explicit counter-position to general purpose providers, particularly for high volume customer support where every 100 milliseconds of lag and every fraction of a cent per minute is felt.
More information
Primary source: Smallest.ai press release, PR Newswire, July 30, 2026
Supplementary source: TechCrunch coverage, July 31, 2026